AquaSmart
Automated underwater sampling for precise weight and length distributions.
- Length & weight estimates
- Population weight distribution
- Biomass metrics
AI & computer vision for aquaculture
Nurture with precision.
Harvest with confidence.
AI vision systems that measure fish inside sea cages and count and weigh juveniles in nurseries — turning what happens under the water into data for feeding, grading and harvest decisions.
Two environments. One vision.
Intelligence across the
production journey.
Two purpose-built systems for two operational realities: measuring fish in open-water sea cages, and counting and weighing juveniles inside existing nursery lines.
AquaVision
Count and weigh juveniles right after the sorter, without slowing the line.
- Fish count
- Individual weights
- Real-time weight histogram
AquaSmart · Sea-cage monitoring
Measure the population
without interrupting it.
Placed directly in the cage, AquaSmart uses AI and advanced vision to assess the size and weight of your stock automatically. This clear picture of growth makes it easier to manage feed and plan your harvest.
- Where
- Inside sea cages, under normal operating conditions
- Input
- Underwater camera footage
- Output
- Length and weight per fish, population weight distribution and biomass metrics
- Used for
- Feeding, grading and harvest planning
How it works
-
Observe
Capture fish in place
Underwater cameras record fish in operational cage conditions.
-
Detect
Find and track fish
Computer vision detects, segments and tracks individual fish across the captured footage.
-
Measure
Estimate size and weight
Calibrated measurements convert each fish’s geometry into size and weight estimates.
-
Aggregate
Build the distribution
Individual estimates become population statistics and growth trends.
See what the system sees
From raw footage to
decision-ready data.
Drag the divider to compare raw underwater footage with AquaSmart’s analysis: every detected fish is segmented and its length estimated, together with the model’s confidence score.
Weight distribution
Measured fish weights from one AquaSmart sample, with a skew-normal fit
Hover over or tap a bar for details.
- Fish measured
- 675
- Fitted mean
- 448.2 g
- Fitted std. deviation
- 208.5 g
- Skew parameter
- 13.94
| Weight range (g) | Number of fish |
|---|
AquaSmart results
From individual fish to
the population picture.
Go beyond simple averages. AquaSmart pairs average weight with full distribution histograms and biomass metrics, so feeding, grading and harvest timing rest on measured data.
- Weight distribution Understand the spread of fish sizes and weights within the population.
- Growth monitoring Compare population structure across different observation periods.
- Production decisions Support feeding, grading and harvest planning with measured data.
AquaVision · Nursery counting & weighing
Inspection that fits
the workflow you already have.
AquaVision combines high-speed machine vision and onboard processing to count fish, measure individual weights and generate real-time weight histograms directly after your sorter — delivering full batch analytics without slowing down the line.
- Where
- Nursery lines, directly after the sorter
- Input
- High-speed images of each passing fish
- Output
- Fish count, individual weights and a real-time weight histogram
- Used for
- Full batch analytics without slowing the line
Where it fits in the line
-
Sort
Keep your current routine
Fish leave the sorter as usual; nothing changes upstream.
-
Capture
Image every passing fish
High-speed machine vision captures each fish as it passes.
-
Analyse
Count and weigh
Onboard processing counts the fish and estimates the weight of each one.
-
Report
Full batch analytics
A real-time weight histogram and a complete batch report, while the line keeps running.
Innovation Lab
Have a problem on the farm?
Bring it to us.
Not every aquaculture challenge has an off-the-shelf answer. If there is something you want to measure, monitor, automate or improve, talk to us. We will explore the problem with you and turn it into a practical computer-vision or automation solution.
Discuss your challenge-
Start with the problem
Tell us what isn’t working.
A difficult measurement, a manual inspection, wasted feed, inconsistent grading, a hidden health issue, or a process you simply wish you could see better.
-
Understand the real environment
We work around the conditions you have.
Turbidity, motion, changing light, depth, biofouling, fish behaviour, hardware limitations, and the way your team actually works on site.
-
Build & validate
Turn the idea into something measurable.
We combine imaging, sensors, machine learning and automation to develop focused solutions — and test whether they deliver useful results in the real aquaculture workflow.
About Abzinnova
Aquaculture improves when uncertainty becomes measurable — and measurements become decisions.
Abzinnova develops computer-vision and automation systems for aquaculture, combining field hardware, edge AI, biological measurement and production workflows to help farms manage stock and production cycles better.
Based at the Azadi Innovation Factory, Tehran
-
Field reality first
Technology is designed around real farms, real water, real fish, and the constraints operators actually face.
-
Evidence-led results
Measurements, models and recommendations are grounded in observable data rather than assumptions.
-
Human-useful output
Complex sensing and AI become information that helps people make better decisions in the field.
Partners
Organisations we work with
Start a conversation
Let’s make the invisible
operational.
Tell us about your farm, nursery, validation need or custom computer-vision challenge. We will route the conversation to the right technical team.
Helpful to include: the species, production stage, your farm or nursery set-up, and what you would like to measure.